Fresh power tubes.......then smoke......

[Paultergeist]

Okay gang, back to that amp I am still trying to fix......

I have basically performed a complete filter cap job on this amp, but i have stopped short of replacing all of the capacitors because doing so is going to be a lot of work. If I have to go that far, I may as well rebuild the entire amp, and consider dropping a different circuit into it as well. As this point in time, I would like to repair this amp to a functional status.

At the rectifier, both of the cap cans, the diodes, and the quieting capcitors were replaced. All (3) main filter capacitors were replaced, along with some of the dropping resistors (all dropping resistors checked). I also replaced the power tubes' cathode by-pass resistor and capacitor.

The amp is still acting up. Whatever the instability issue is (noise, static, random sputters), it is not the rectifier or filter caps, as they have all been replaced. It is not the tubes either.

While checking pin voltages, I noticed something odd at the phase inverter (PI) 12AU7 tube: Voltages

Pin 1: 150
Pin 2: 0
Pin 3: 5

Pin 6: 150
Pin 7: 23 (this voltage wandered a bit)
Pin 8: here is where it got odd. Several times I put the DMM probe on pin 8 and saw "80 volts," only to then have that reading drop down after a second or two and go down as low as zero. The reading seemed most stable, however, at 18 volts.

I confess that I do not understand this type of PI very well, but I am looking for help. I am pretty much out of ideas on what is going on with this amp, but if there is something funny going on with the power supply to that PI tube, maybe I can sort this out. The resistors immediately nearby seemed about right, but my thinking is that maybe I have a leaking cap nearby?

Any thought are much appreciated.

The circuit:

http://www.larryjohnmcnally.com/Teisco/ckmt20_sch.pdf

[Paultergeist]

I am going to re-fine my question a bit:

I now think I have narrowed this down to a cathodyne type phase inverter (PI). here are the voltages:

Plate (Pin 6): 150
Grid (Pin 7): 23 (this voltage seemed to wander a bit)
Cathode (Pin : here is where it got odd. Several times I put the DMM probe on pin 8 and saw "80 volts," only to then have that reading drop down after a second or two and go down as low as zero. The reading seemed most stable, however, at 18 volts.

I believe I have instability at the Cathode / Grid. I question whether there should be much of any voltage potential at the grid if the amp is "ON" but naot making sound (quiescent).

I am secondly trying understand what the voltage should be at a cathodyne PI cathode? And how the voltage is supplied / stabilized at the cathode?

[Mescalero]

Check the coupling caps and grid-leak resistors at the phase splitter.

[Mescalero]

D'oh! Duplicate post.

That is a grid-leak bias configuration. The cathode will be above ground due to I*R drop on the 30 k Ohm resistor and the grid will be slightly (~1.5V?) negative when measured relative to the cathode.

[Paultergeist]

Mescalero,

Thanks very much for the helpful replies.

Question: am I correct in referring to this driver as a "cathodyne" (PI)?

As soon as I get home from work tonight, I will double-check the voltages at the suspect triode, and measure the values across the adjacent resistors, then report back (here).

[Mescalero]

http://www.valvewizard.co.uk/cathodyne.html
It is, and here's a good description of it:

[Paultergeist]

Thanks Mescalero,

Indeed -- as you suggested -- the grid measured -1.6 V referenced to the cathode.

The plate is at 160V, the Cathode is at 85V.

I suspect that I may not have had a very secure contact with the DMM probes when I had measured previously. I was so hoping to home-in on the problem with this amp, but it really feels to me like I am truly in over my head.

The amp has quite a number of (original?) paper-in-oil signal and coupling caps, and my knee-jerk suspicion is that one of them may have gone bad, though I can see no clear sign of leaking. Because of this point-to-point "tag strip" construction method, it is proving to be rather destructive to the amp to de-solder things, as multiple component leads typically share a single eyelet, and the metal has grown rather brittle over all this time. In other words, I am reluctant to try and just replace all the signal/coupling caps.

I am open to any ideas, but may need to give up the ghost on this one.....?

Thanks very much for the help thus far. I do appreciate it

[Mescalero]

Cutting the old cap free first can help minimize the strain on lugs. Another way to do it would be to cut the old cap free but leave some of its leads behind. Then the new cap's leads are spiraled around a tiny screw driver or similar tool to form two tiny cylinders and the old leads are then each captured within a spiral and soldered. This certainly does not look stock, but, it can save a lot of hassle and mitigates damage potential.

[Mescalero]

www.renovatedradios.com/articlePages/H5 ... 0Guide.pdf

A very groovy man named John "badrestorer" has taken a great deal of time to generate tutorials on radio repair. On page 50 is a bit on re-stuffing can caps. The spiral that I talk about is on page 54. The whole document is right fair reading. Enjoy.

[Paultergeist]

Mescalero,

It is most kind of you to assit me. The document from "bad restorer" was fascinating. Thank you. It is somewhat re-assuring to observe his technique for replacing components using a twist on the new leads to slip over existing leads. I may be able to use that in at least a few of the instances in this amp.

If I have not already used up all of my lifelines, I would value your opinion.

This amp is about 50 years old, and had original everything, the issue is random noise -- snap, crackle, pop, sounds like ocean waves -- but it is intermittent. Sometimes the amp sounds pretty decent, and the guitar signal is coming through. Through process of elimination, I do not believe it to be any of the tubes (although I do not have a reverb tube to swap), and it cannot be the power supply filtering rail. So....my question.......before I gamble on more component costs......how probable is this random, intermittent noise likely to be one of those original signal /coupling caps?

I realize this is pure speculation, but I am trying to gauge if I am just wasting money of it I am likely to actually be getting close to finding the culprit?

Thanks again, and I'll understand if you would rather not speculate.....

[RWood]

Is it true PTP or does it have an eyelet board? It sounds a lot like Tweed disease on those lousy black Fender boards when they become conductive.

[Mescalero]

With one hand in your pocket, try wiggling the tubes when it occurs to see if it may be caused by oxidation of the pins/sockets. If so, a little (not a lot) of DeOxit is your friend.
I agree that it may be contamination causing errant behavior. If there is significant (i.e.: an oily residue or tobacco smoke film) contamination, the old-timer's trick was a good wash with naptha (lighter fluid). Sockets and boards alike would be lovingly brushed down (outside!) and then left to dry in the sun.

[Paultergeist]

Is it true PTP or does it have an eyelet board? It sounds a lot like Tweed disease on those lousy black Fender boards when they become conductive.

Thanks for the input......this is true PTP -- the first of such kind I have ever worked on (not that my experience is all that extensive!) If you are curious, I can try and post a photo(?)

Essentially, there is a single row of tags strips down the longitudinal axis of the chassis -- this row of tag strips bridges the space between the row of tube sockets and the front panel (pots) of the amp. Almost all of the components float between the tag strip and the front panel, or the tag strip and the tube sockets. There is no eyelet or turret board.

The issue I am encountering with this construction method is that -- for instance -- a single connection point may serve multiple components (and of course, the component I want to replace is usually buried beneath other components). Those pre-amp tube sockets look like *clumps* of resistors/capacitors from inside the chassis!

I am familiar with the semi-conductivity issue seem with the old black fiber eyelet boards -- not the case here. Thanks for chiming in!

[Paultergeist]

With one hand in your pocket, try wiggling the tubes when it occurs to see if it may be caused by oxidation of the pins/sockets. If so, a little (not a lot) of DeOxit is your friend.

I'll try this. I will point out that I have -- individually -- swapped out each tube in the amp except for the rather obscure reverb driver -- the problem was unchanged during this attempt. I will try a more focused exploration of the tube sockets themselves. If it is the sockets, then I should be able to re-create the issue on demand, instead of the random nature I am now experiencing.

I agree that it may be contamination causing errant behavior. If there is significant (i.e.: an oily residue or tobacco smoke film) contamination, the old-timer's trick was a good wash with naptha (lighter fluid). Sockets and boards alike would be lovingly brushed down (outside!) and then left to dry in the sun.

Well, I have read about paper-in-oil caps "leaking," so I have been looking for residue coming out of these caps -- but I just do not see any significant signs of oil or film.

???

[Phil_S]

[I have read about paper-in-oil caps "leaking," so I have been looking for residue coming out of these caps -- but I just do not see any significant signs of oil or film.

Well, yes, we could possibly be discussing physical leaking of the oil. But, no...leakage is an electronics term. Caps are supposed to block DC and pass AC voltage. Leakage refers to cap failure when it "leaks" DC because the insulating medium between the layers has broken down. At this point the cap is said to be leaky and it is no longer doing its intended job.

If your signal caps are passing DC, they need to be replaced.